O R I G I N A L A R T I C L E S UDC: 613.86-053.8::616.89-07]:[314.04:616-058 DOI: 10.2298/VSP1209747M
Assessment of mental health in adults of the northern part of the city
of Kosovska Mitrovica
Procena mentalnog zdravlja odraslih stanovnika severne Kosovske Mitrovice
Momþilo Mirkoviü*, Snežana Simiü†, Goran Trajkoviü‡
*Institute of Social Medicine, School of Medicine, University in Pristina, Kosovska Mitrovica, Serbia; †Institute of Social Medicine, ‡Institute of Medical Statistics, School
of Medicine, University in Belgrade, Belgrade, Serbia
Abstract
Background/Aim. Mental health disorders lead to disorder of effective functioning of people and deterioration of quality of life. Early detection of individuals at risk of mental health disorders is extremely important from the aspect of mental health disorders prevention. The aim of the research was to determine the frequency of mental health problems among adult residents of northern Kosovska Mitrovica and to exam-ine the association between frequency of mental health prob-lems and socio-demographic and other characteristics of the population obtained by the questionnaire. Methods. The cross-sectional study on the representative sample of adult residents of northern Kosovska Mitrovica was performed in October 2009. To obtain information about the characteristics of mental health the Goldberg’s General Health Questionnaire (GHQ-28) was used. For performing survey at site the method of rapid epidemiological assessment was chosen. Sta-tistical analysis included the methods of descriptive statistics, multivariate regression analysis and calculation of the Cron-bach’s alpha coefficient of internal consistency of the ques-tionnaire. Results. Mental health problems (total score) were present in almost half of the respondents (49.2%). Psychoso-matic problems were present in more than half of the respon-dents (55.4%), while anxiety and insomnia were present in al-most half of the respondents (49.2%). Social dysfunction had more than three fifths of the respondents (63.1%) and depres-sion more than a quarter of the respondents (28.5%). More positive responses in the questionnaire were statistically sig-nificantly associated with older age, poor financial situation, abuse and assessing of the current political-security situation as high risk. The value of Cronbach’s alpha coefficient was 0.705. Conclusions. Almost half of the respondents (49.2%) of North Kosovska Mitrovica had mental health problems. Mental health problems were associated with older age, poor financial situation, abuse and considering the current political-security situation as high-risk factor.
Key words:
mental heath; risk factors; risk assessment; adult; serbia.
Apstrakt
Uvod/Cilj. Poremeýaji mentalnog zdravlja dovode do po-remeýaja efektivnog funkcionisanja ljudi i znaÿajno utiÿu na pogoršanje kvaliteta života. Rano otkrivanje osoba se rizi-kom od obolevanja od mentalnih poremeýaja izuzetno je znaÿajno sa stanovišta prevencije mentalnih problema. Cilj istraživanja bio je da se utvrdi uÿestalost problema sa men-talnim zdravljem meĀu odraslim stanovnicima severne Ko-sovske Mitrovice i ispita povezanost uÿestalosti problema sa mentalnim zdravljem i socijalno-demografskih i drugih ka-rakteristika stanovnika dobijenih upitnikom. Metode. Istra-živanje je uraĀeno kao studija preseka na reprezentativnom uzorku odraslog stanovništva severne Kosovske Mitrovice u oktobru 2009. godine. Za dobijanje informacija o karakteri-stikama mentalnog zdravlja korišýen je Goldbergov Upitnik za opšte zdravstveno stanje. Za anketiranje na terenu izab-ran je metod brze epidemiološke procene. Statistiÿka analiza ukljuÿila je metode deskriptivne statistike, multivarijantnu regresionu analizu i izraÿunavanje Kronbahovog alfa koefi-cijenta interne konzistencije upitnika. Rezultati. Probleme sa mentalnim zdravljem (ukupan rezultat) imala je skoro polovina ispitanika (49,2%). Psihosomatske probleme imalo je nešto više od polovine ispitanika (55,4%). Anksioznost i insomniju imala je skoro polovina ispitanika (49,2%). Soci-jalnu disfunkciju imalo je više od tri petine (63,1%) a depre-siju nešto više od ÿetvrtine ispitanika (28,5%). Veýi broj po-zitivnih odgovora u upitniku statistiÿki je znaÿajno povezan sa starijim životnim dobom, lošijim materijalnim stanjem, zlostavljanjem i procenom trenutne politiÿko-bezbednosne situacije kao visoko riziÿne. Vrednost Kronbahovog alfa ko-eficijenta iznosila je 0,705. Zakljuÿak. Skoro polovina (49,2%) stanovnika severne Kosovske Mitrovice imala je problema sa mentalnim zdravljem. Problemi sa mentalnim zdravljem povezani su sa starijim životnim dobom, lošim materijalnim stanjem, zlostavljanjem i procenom trenutne politiÿka-bezbednosne situacije kao fakotrom visokog rizika.
Kljuÿne reÿi:
Introduction
According to the definition of the World Health Or-ganization, which is an integral part of the Constitution of this organization 1, mental health is a component of health. The World Health Organization points out that mental health is not merely the absence of mental illness, but it is "a state of well-being in which the individual realizes his or her own abilities, can cope with the normal stresses of life, can work productively and fruitfully, and is able to make a contribu-tion to his or her community" 2. Mental health disorders lead to disorders of functioning and significantly affect the quality of life. In addition, mental health disorders are associated with physical health disorders, particularly with chronic dis-eases from which many people die and suffer, such as car-diovascular disease, diabetes, and etc. 3, 4.
Early detection of people at risk of mental disorders de-veloping is extremely important in prevention of mental disor-ders 5. So far, a few questionnaires have been created that can help to detect people at risk of mental disorders developing, but the Goldberg`s General Health Questionnaire is one of the most commonly used for this purpose 6. This questionnaire has been used in various investigations. This questionnaire has been used to examine mental health of users of primary health care 7, workpeople 8, patients with ischemic heart disease and patients with gastrostoma 9, 10, and to investigate mental health of survivors of the tsunami in Indonesia 11.
Within the last decade of the 20th century, Serbia faced many stressful events, such as disintegration of the former state, the United Nations economic sanctions that lasted three and a half years, conflicts in Kosovo, the bombing by NATO in 1999. All this had an effect on mental health of the popu-lation, so that the prevalence of mental disorders increased by 13.5% in the period 1999–2002 12. In a study on burden of disease and injury in Serbia in 2002 depression is ranked fourth (third for women, and sixth for men) 13 . After the bombing, in June 1999, the Autonomous Province of Kosovo and Metohia came under UN administration 14, and stressful events continued, especially for non-Albanian population. More than half of Serbs fled the Province, and those who remained were subjected to daily attacks. Many of them were kidnapped or killed 15. Living in such an unstable political and security area affects mental health, but the data on this component of health are deficient. In September and October 1999 Salama et al. 16 were carried out the study on mental health of the population of Gnjilane and Priština using the Goldberg`s a questionnaire. In 2006 under the Health Survey the residents of northern Kosovska Mitrovica were investi-gated and some aspects of their mental health17. According to this survey, more than half of the respondents had problems with some of the aspects of mental health, urging to conduct a survey with on instrument adapted to assess mental health.
The aim of this research was to determine the preva-lence of mental health problems in adult residents of North Kosovska Mitrovica and to study the relationship between mental health problems and the basic characteristics of adult residents of northern Kosovska Mitrovica obtained by the questionnaire (gender, age, marital status, number of family
members, education, employment, self-rated financial situa-tion, whether a person is displaced, the existence of mental disorders in the family, social support, abuse, assessment of current and future political and security situation).
We hypothesized that mental health was worse in fe-males, adults, residents with lower education, unemploy-ments, poor, displaced persons, abused and with poor social support. Also, we hypothesized that mental health could be worse in those people who take the political-security situa-tion as high-risk factor.
Methods
This cross-sectional study included the representative sample of adults in northern Kosovska Mitrovica, and was performed in October 2009.
The sample was selected by the list of polling units with the number of votes for each constituency and the list of streets that includes voting record. These data are the part of electoral roll for the parliamentary elections in the Republic of Serbia in 2008. A total number of voters in northern Kosovska Mitrovica was 17,876 residents. The required sample size for assessing the prevalence of mental disorders, with the assumption of their frequency in a population of 20%, with an accuracy of 0.7 (7%) and the confidence level of 95% was 126.
To obtain information about socio-demographic char-acteristics of respondents we used the questionnaire with questions about gender, age, education, occupation, self-rated financial situation, marital status, the presence of men-tal disorders in the family, social support, abuse and, consid-ering the characteristics of the territory in which they lived, whether or not displaced and if take the current and future political and security situation as risky.
To obtain information about the characteristics of men-tal health we used the Goldberg`s General Health Question-naire (GHQ-28) 7. This questionnaire allows to estimate the prevalence of mental health problems and psychological dis-tress in the target population 18.
weeks as compared to normal. The classical way of scoring was as follows: the answers 1 and 2 were scored by 0 (zero), and the answers 3 and 4 scored by 1 (one). This scoring method seemed to be sensitive to temporary conditions and allowed detection of mental health deterioration 18. The scoring in this way did not reveal long-term chronic prob-lems with mental health 19. To overcome this deficiency, a new scoring was proposed in 1985, that only the first answer was scored with 0, and the others with 1 19.Since the tested territory political and security situation deteriorated for many years, we applied a new, modified scoring system proposed in 1985. The author of the questionnaire in 1998 suggested the best threshold to be arithmetic mean of all points of re-spondents 20, which was used in this research.
Because of the source of information, for the sample of respondentsat site the method of rapid epidemiological as-sessment was chosen 21. For each constituency the number of respondents was determinedand divided into the streets that made up the constituency. The number of respondents was pre-determined for each street to be interviewed and started from the first building or house on the right side of the start. If it was a building, we first surveyed the people from the first flat and then from every second. One flat was skipped. In case of not cooperating in the survey we skipped apart-ment, according to the ordinal numbers. In case of a house at the beginning of the street, the survey was performed in the first house and then in every second house. The survey in one street was considered finished when included as many respondents as possible. The surveyed people were older than 18 years. Due to the nature of the issues, the author in-sisted that the respondents complete the questionnaire inde-pendently, and only when they insisted the interviewer read questions and recorded answers, which was significantly less frequent, mainly for the elderly respondents. In this way, we surveyed 130 respondents in 40 households. In five house-holds respondents refused to participate in the survey, which means that the rate of response was 89.0%.
The methods of statistical analysis included descriptive statistics and multivariate regression analysis (using the pro-cedure step by step). To determine the reliability of research instrument we determined the Cronbach alpha coefficient of internal consistency of the questionnaire.
Results
The study included 130 respondents. More than half (56.2%) of them were female. The average age was 45.5 ± 17.3 years. Half of the respondents had secondary
level of education, and most of them were married (58.5%). More than half of the respondents (56.9%) were employed. Most of them (42.3%) described their financial condition as middle (Table 1).
Table 1 Demographics and socio-economics characteristics of
respondents
Characteristics n (%)
Gender
Male 57 (43.8)
Female 73 (56.2)
Age groups
18–24 18 (13.8)
25–34 22 (16.9)
35–44 24 (18.5)
45–54 26 (20)
55–64 20 (15.4)
65+ 20 (15.4)
Education
With not school 3 (2.3)
Incomplete primary school 3 (2.3) Complete primary school 10 (7.7)
High school 65 (50)
College 25 (19.2)
Faculty 24 (18.5)
Marital status
Married 76 (58.5)
Cohabitation 2 (1.5)
Single 35 (26.9)
Divorced 4 (3.1)
Widowed 13 (10)
Employment
Employed 74 (56.9)
Self-employed 3 (2.3)
Retired 24 (18.5)
Housewife 3 (2.3)
Student 19 (14.6)
Unemployed 7 (5.4)
Unable to work 0 (0)
Financial situation
Very bad 1 (0.8)
Bad 6 (4.6)
Middle 55 (42.3)
Good 53 (40.8)
Very good 15 (11.5)
Table 2 shows the values of arithmetic means of the re-sults obtained by the Goldberg`s General Health Questionnaire – 28 (GHQ-28), as the threshold for determining whether a re-spondent is in good or poor mental health as well as the pro-portion of respondents who scored below or above the mean value. Thus, the psychosomatic problems of all respondents
Table 2 The results obtained by the Goldberg's General Health Questionaire-28 (GHQ-28)
< ʉ > ʉ Different aspects of mental health ʉ sd min max
n (%) n (%)
Psychosomatic problems 4.61 2.12 0 7 58 (44.6) 72 (55.4)
Anxiety and insomnia 4.18 2.32 0 7 66 (50.8) 64 (49.2)
Social dysfunction 5.52 1.76 1 7 48 (36.9) 82 (63.1)
Depression 1.22 1.88 0 7 93 (71.5) 37 (28.5)
with 5 or more positive answers were classified as positive, and they were more than half of the respondents (55.4%). The same result of arithmetic mean was for anxiety and insomnia, that were present in almost half of the respondents (49.2%). There were 6 positive responses for social dysfunction thresh-old, present in more than three-fifths (63.1%) of the respo-dents. The threshold for depression was 2, which was marked by more than a quarter of respondents (28.5%). The threshold for the total score was 16 positive responses, which was given by almost half of the respondents (49.2%).
Table 3 shows the results of multiple linear regression for total score related to mental health. There is a highly sta-tistically significant positive correlation between mental
health problems and age, which means that these problems are more frequent in the elderly. Problems with mental health were significantly negatively associated with financial state and abuse, showing that mental health problems are more frequent in patients with poor financial situation and abuse. A statistically significant positive correlation was found be-tween mental health problems and assessment of the current political and security situation showing that those who con-sider it as high risk are in worse mental health.
Table 4 presents the results of multiple linear regres-sion for psychosomatic problems, anxiety and insomnia, social functioning and depression. Psychosomatic problems were significantly positively associated with age, social support and assessment of future political and security situation and negatively with their financial situation, im-plying that these problems were more frequent in the eld-erly, those without social support, who consider future po-litical and security situation as high risk and those who were in worse financial situation. Social functioning was significantly positively correlated with age, implying that older people had more frequent problems with social func-tioning. Anxiety and insomnia were significantly positively
associated with age and assessment of the current political and security situation, showing they were more frequent in the elderly and those who considered current political and security situation as high risk. Depression was statistically significantly positively associated with age and marital status, and negatively with their financial situation, that was more frequent in the elderly, widowed and divorced, those with poor financial condition and in people suffering from abuse.
Table 3 Multiple linear regression model (step-by-step) for mental health in relation to variables that are identified as predictors
Unstandardized coefficient
Standardized coefficient Explanatory variables (predictors)
b SE ȕ
t p
Age 0.12 0.02 0.338 4.79 0.000
Financial situation -1.48 0.53 -0.197 -2.79 0.006
Abuse -2.96 1.22 -0.16 -2.42 0.017
Assessment of the current political-security situation 2.24 0.48 0.333 4.66 0.000
Constant 17.76 4.63 3.84 0.000
Table 4 Multiple linear regression model (step-by-step) for different aspects of mental health
in relation to variables identified as predictors
Unstandardized coefficient Standardized coefficient Explanatory variables (predictors)
b SE ȕ
t p
Psychosomatic disorders
Age 0.03 0.009 0.28 3.62 0.000
Financial situation -0.058 0.208 -0.21 -2.77 0.007
Social support 2.29 1.027 0.16 2.24 0.027
Assessment of the future
political-security situation 0.77 0.183 0.32 4.19 0.000
Constant 5.33 1.592 3.35 0.001
Social functioning
Age 0.04 0.01 0.342 4.12 0.000
Constant 3.94 0.41 9.62 0.000
Anxiety and insomnia
Age 0.04 0.01 0.26 3.4 0.001
Assessment of the current
political-security situation 1.15 0.2 0.436 5.69 0.000
Constant -0.74 0.65 -1.14 0.256
Depresion
Age 0.02 0.01 0.208 2.74 0.007
Marital status 0.23 0.1 0.17 2.29 0.024
Financial situation -0.87 0.18 -0.363 -4.78 0.000
Abuse -1.5 0.43 -0.256 -3.43 0.001
The value of Cronbah alpha coefficient of internal con-sistency of the questionnaire was 0.705 (95% confidence in-terval: 0.612 to 0.780), indicating satisfactory reliability of the questionnaire, as the values higher than 0.7 are inter-preted as acceptable 22.
Discussion
This research examined the prevalence of mental health problems of residents in northern Kosovska Mitrovica, and the association of mental health problems with demographic, socio-economic and other characteristics of the population, resulting from the characteristics of the territory. The applied questionnaire – GHQ-28, for mental health assessment in previous research showed to have acceptable reliability and validity for use for adult population in different countries. It has been used as a screening test by the World Health Or-ganization in one of multicenter studies 6. The applied scor-ing results proved satisfactory for the detection of chronic problems with mental health 23.
The results showed that the average score of positive an-swers in the questionnaire was 15.54, taken as the threshold for determining the presence of mental health problems, so that each respondent with 16 or more positive answers was in-dicated as positive. These respondents were nearly half (49.2%). In a similar research for the rankings, which was done for the region of Montreal in Canada, the limit value was 11.45, a positive was 44.5% respondents 5. This shows that all respondents in northern Kosovska Mitrovica had poorer men-tal health as a whole, and in addition, higher proportion of re-spondents had mental health problems. Salama et al. 16 have come to the conclusion that the average score of positive an-swers for the Serbs from Kosovo and Metohija is greater than all the population in Europe or in other parts of the world. This result coincides with the results of studies that have shown deterioration in mental health of residents living in areas af-fected by war and unstable political-security situation 24–26.
The study also found that mental health problems were related to older age, poor financial situation, abuse and assess-ment of the current political-security situation as high risk. This result coincides with the evidence from literature 2, 27, which states that these characteristics are among the most important determinants of mental health. A connection between mental health problems and an assessment of the current political-security situation as high risk is particularly interesting, indi-cating that the fear of evidently deteriorated political and se-curity situation affect mental health so to cause
psychologi-cal distress. This result makes the research itself justified, given that the data about mental health of residents of this territory, especially on its determinants, is deficient. In addi-tion, the Serbian Government in its Strategy for the Devel-opment Care for Mental Health points out that it is necessary to support research in all areas of mental health in our com-munity and that epidemiological studies on this subject are extremely rare 28. Determination of residents with mental health problems, but not to the extend of illness, and to de-fine risk groups is very important in prevention of mental disorders. Preventive strategies are maximally effective if applied before the onset of the disorder and if directed to the groups at higher risk of suffering 29. By establishing that, nearly half the population of the area have mental health problems highlights the importance of this problem, imply-ing the need to implement preventive strategies.
Possible limitations of this study come from the type of study, which examines the relationship between variables in a defined point in time, so it could lead to call in question a variables causality, since there is no information on their re-lationship in the long time. The average score of positive an-swers used as a threshold or borderline between good and poor mental health was higher than in other studies. Increas-ing the threshold in the interpretation of the results leads to a decrease in sensitivity and increase in specificity of instru-ment 30, which also could be a limitation of the study. How-ever, data about sensitivity and specificity in this case are not known, since we do not use another instrument, so the results could not be compared. The hypothesis of the research is partly confirmed, a restriction may be the sample size, be-cause the biger sample allows to detect more discrete asso-ciation as statistically significant. Finally, the data could be generalized only to the area of northern Kosovska Mitrovica, with respect to the method used for sampling.
Conclusion
The obtained results show that almost half (49.2%) of the residents of northern Kosovska Mitrovica have mental health problems. Mental health problems are associated with older age, poor financial condition, abuse and taking the cur-rent political-security situation as high risk. Theses results partially confirm the hypothesis of the study, but determining the frequency of mental health problems and risk groups is important for planning and implementing prevention strate-gies for preservation, enhancement and improvement of mental health.
R E F E R E N C E S
1. World Health Organization. Constitution os the World Health Organization. In: World Health Organization, editor.Basic documents. 45th ed. Geneva: WHO; 2006.
2. World Health Organization.. Promoting Mental Health – A Re-port of the World Health Organization. Geneva: WHO; 2005.
3. Pinto-Meza A, Fernández A, Fullana MA, Haro JM, Palao D, Lu-ciano JV, et al. Impact of mental disorders and chronic
physi-cal conditions in health-related quality of life among primary care patients: results from an epidemiological study. Qual Life Res 2009; 18(8): 1011î8.
GHQ in a primary care setting. Soc Psychiatry Psychiatr Epi-demiol 2004; 39(3): 235î43.
6. Goldberg DP, Gater R, Sartorius N, Ustun TB, Piccinelli M, Gureje O, et al. The validity of two versions of the GHQ in the WHO study of mental illness in general health care. Psychol Med 1997; 27(1): 191î7.
7. de la Revilla Ahumada L, de los Ríos Alvarez AM, Luna del Castillo JD. Use of the Goldberg General Health Questionnaire (GHQ-28) to detect psychosocial problems in the family phy-sician's office. Aten Primaria 2004; 33(8): 417î22; discussion 423î5. (Spanish)
8. Moreno-Abril O, Luna-del-Castillo Jde D, Fernández-Molina C, Ju-rado D, Gurpegui M, Lardelli-Claret P, et al. Factors associated with psychiatric morbidity in Spanish schoolteachers. Occup Med (Lond) 2007; 57(3): 194î202.
9. Failde I, Ramos I, Fernandez-Palacín F. Comparison between the GHQ-28 and SF-36 (MH 1-5) for the assessment of the mental health in patients with ischaemic heart disease. Eur J Epidemiol 2000; 16(4): 311î6.
10.Mahjoubi B, Mohammadsadeghi H, Mohammadipour M, Mirzaei R, Moini R. Evaluation of psychiatric illness in Iranian stoma pa-tients. J Psychosom Res 2009; 66(3): 249î53.
11.Heir T, Weisaeth L. Acute disaster exposure and mental health complaints of Norwegian tsunami survivors six months post disaster. Psychiatry 2008; 71(3): 266î76.
12.Lecic Tosevski D, Pejovic Milovancevic M, Popovic Deusic S. Reform of mental health care in Serbia: ten steps plus one. World Psychiatry 2007; 6(2): 115î7.
13.Atanaskoviý-Markoviý Z, Bjegoviý V, Jankoviý S, Kocev N, Laaser U, Marinkoviý J, et al. The Burden of Disease and Injury in Ser-bia. Belgrade: Ministry of Health; 2003. (Serbian)
14. UN. Resolution 1244. New York: Security Council; 1999 [cited 2009August 26]. Available from: http://daccess-ods.un.org/TMP/2961369.html
15.The Government of the Republic of Serbia. Strategy for sustainable survival and returt to Kosovo and Metohija. Belgrade: The Government of the Republic of Serbia; 2009. (Serbian) 16.Salama P, Spiegel P, Van Dyke M, Phelps L, Wilkinson C. Mental
health and nutritional status among the adult Serbian minority in Kosovo. JAMA 2000; 284(5): 578î84.
17.Mirkoviý M. Assessment of health status of residents in socio-economic and political-security vulnerable territory [thesis]. Belgrade: School of Medicine; 2008. (Serbian)
18.Golderberg D, Williams P. A user's guide to the General Health questionnaire. Windsor, UK: NFER-Nelson; 1988.
19.Goodchild ME, Duncan-Jones P. Chronicity and the General Health Questionnaire. Br J Psychiatry 1985; 146: 55î61. 20.Goldberg DP, Oldehinkel T, Ormel J. Why GHQ threshold varies
from one place to another. Psychol Med 1998; 28(4): 915î21. 21.MacIntyre K. Rapid assessment and sample surveys: trade-offs in precision and cost. Health Policy Plan 1999; 14(4): 363î73. 22.McHorney CA, Ware JE Jr, Lu JF, Sherbourne CD. The MOS 36-item Short-Form Health Survey (SF-36): III. Tests of data quality, scaling assumptions, and reliability across diverse pa-tient groups. Med Care 1994; 32(1): 40î66.
23.Huppert FA, Gore M, Elliott BJ. The value of an improved scoring system (CGHQ) for the General Health Question-naire in a representative community sample. Psychol Med 1988; 18(4): 1001î6.
24.Kozaric-Kovacic D, Folnegovic-Smalc V, Skrinjaric J, Szajnberg NM, Marusic A. Rape, torture, and traumatization of Bosnian and Croatian women: psychological sequelae. Am J Orthopsy-chiatry 1995; 65(3): 428î33.
25.Mollica RF, McInnes K, Sarajliý N, Lavelle J, Sarajliý I, Massagli MP. Disability associated with psychiatric comorbidity and health status in Bosnian refugees living in Croatia. JAMA 1999; 282(5): 433î9.
26.Coyne JC, Kagee A. Mental health among Bosnian refugees. JAMA 2000; 283(1): 55.
27.World Health Organization. Mental Health: New Udestanding, New Hope. In: World Health Organization, editor. World Health Report 2001. Geneva: World Health Organization; 2001.
28.The Government of the Republic of Serbia. Strategy for development of mental health care. Belgrade: The Government of the Re-public of Serbia; 2007. (Serbian)
29.World Health Organization.Prevention and promotion in mental health. Geneva:World Health Organization; 2002.
30.Andersen HS, Sestoft D, Lillebaek T, Gabrielsen G, Hemmingsen R. Validity of the General Health Questionnaire (GHQ-28) in a prison population: data from a randomized sample of prison-ers on remand. Int J Law Psychiatry 2002; 25(6): 573î80.
